mejutoco

33 ms·

481.
by mejutoco 3y ago
I agree with you with it on the scale example. Thanks for the detailed explanation. Where we disagree fundamentally is that you are finding that the analogy is wrong. But wrong for that? For the scale example it is wrong. For the aerodynami
482.
by mejutoco 3y ago
Google might wait 6 months until they rank your site higher (same as you mention some old websites). I think for a new website it would take a while to rank properly. This matches my experience with own domains too.
483.
by mejutoco 3y ago
Pebble/pebbles? As in those stones on the bottom of the river polished by the water or sand.
484.
by mejutoco 3y ago
I agree with you and the examples you mention. At the same time I am not disappointed that the model/analogy does not work in all cases. That does not mean it is not illuminating in others. Following your example (kind of, with planes)
485.
by mejutoco 3y ago
I think there are things in common and others not. For analogies, like models, I guess there is no right or wrong, only useful. Some of your examples indeed do not work with the analogy. Others might. For instance, that daughter who is 16 m
486.
by mejutoco 3y ago
I like Pydantic, but it is true it broke too many things. It is not the same, but typeguard has runtime typechecks with the @typechecked function annotation. Mypy checks help as well.
487.
by mejutoco 3y ago
Like in the little prince, where it is stated that adults always ask questions about numbers and quantities :)
488.
by mejutoco 3y ago
I ask it about creating a conversation in Polish with English translations about an encounter between two neighbours walking their dogs.
489.
by mejutoco 3y ago
In Addition to requiring a credit card, you could think of an invite-only system to have a chain of users that vouch for other users.
490.
by mejutoco 3y ago
Obsidian also uses markdown files, so you could have stayed with it and backup the folders normally, if you do not want to pay for the syncing.
491.
by mejutoco 3y ago
But the cloud would cost apple money, while if ran in your own device they can sell you the hardware for a healthy margin, and market privacy. I think a single useful model on device or many tiny ones are feasible. Even today, the photos ap
492.
by mejutoco 3y ago
Since we are on this topic. Yes, maybe 10 years. And since when does _one_ mattress work for a family of 4?
493.
by mejutoco 3y ago
On iOS you can scroll diagonally from top right corner and one more click to disable it. It is not permanent and comes back when disabling airplane mode, in addition to updates too.
494.
by mejutoco 3y ago
Thanks for this comment. I am a big fan of Böll and I still keep discovering interesting things about him.
495.
by mejutoco 3y ago
> (1) What's the value add of the LLM here? Does it substantially reduce the number of evaluations necessary to converge? If so, how? I thought that stochastic (random) gradient descent and LLM were converging much quicker than gene
496.
by mejutoco 3y ago
I wonder how someone will integrate symbolic reasoning with LLMs, or if it will be possible.
497.
by mejutoco 3y ago
PWA features could be a similar situation. They are also not a standard.
498.
by mejutoco 3y ago
This matches some vision of the early Internet and agents (having programs browsing the internet for you to do useful tasks). I thought it was Douglas Engelbart but I cannot find anything about agents. Maybe it was some other technology exp
499.
by mejutoco 3y ago
> If the goal is openness, i.e. why GitHub has the prestige it does, allying itself with open software development goals I guess that is why Microsoft acquired them.
500.
by mejutoco 3y ago
> pollution (from materials and from transport) We agree
501.
by mejutoco 3y ago
> But ultimately every company that has a platform for software development is going to charge you: This is why the web won, IMO.
502.
by mejutoco 3y ago
I think, as a thought exercise, one can make the same argument about plastic. Even though it is cheap and there is demand it does not mean that it does not have also bad consequences (non biodegradable in that case). In the fast fashion cas
503.
by mejutoco 3y ago
DRF is very dynamic. While useful, it does too much magic for my taste. Pydantic pure data models, combined with typeguard annotations do it for me. Plain functions with defined input output types. It helps validate all the assumptions and
504.
by mejutoco 3y ago
> AI doesn't slack when you're not looking at it. While humans do. I have had a different experience. Some humans slack while nobody is looking; some humans slack even if you look; some humans do not slack. Alternatively, as lo
505.
by mejutoco 3y ago
> but also office space, amenities, liabilities. People (probably not the same people) are still resisting this with remote work, but get all excited when it is about AI. Humans are strange.
506.
by mejutoco 3y ago
This has happened with books already and nothing (arguably) broke.
507.
by mejutoco 3y ago
Any will do, so I do not want to nitpick too much. I want to point out that most people are using pytorch these days. If you use the book that uses Keras it is also fine, but I would not start with Tensorflow today.
508.
by mejutoco 3y ago
A very simple one I did is a clone of the iPhone calculator as a PWA so I can use it in the iPad. It is plain Typescript without a framework, can be added to the home screen (PWA) and work offline. https://getcalculator.app/
509.
by mejutoco 3y ago
Long inheritance chains solidify and make it difficult to change code. Same way a platypus is stuck bring a mammal although it lays eggs.
510.
by mejutoco 3y ago
I wonder how AI used "for good" would look like. To take all the disinformation online and point out the contradictions and things that do not make sense in a clear way. Of course, it remains to be seen if people would be convince

More ›